[daip] task suggestions

[Eric Greisen]

David A. Boboltz writes:

 > Task:  SAD
 > 
 > Right now, I run SAD in a loop over channels.  I use IMSTAT to make an
 > estimate of the off-source noise in the corner of the image, then I take
 > this estimate and multiply it (maybe by 8 to 10) and use this to set CPARM
 > and DPARM.   The SAD now seems to provide a better estimate of
 > the off-source noise than my crude use of IMSTAT, and uses this to
 > set CUTOFF (which seems to be permanently set at 3*off-source noise).
 > It would be nice if either the off-source noise estimator could be available
 > in a separate task (maybe it is and I haven't found it yet), so that the 
 > estimate
 > could be fed to SAD to set CPARM and DPARM, or if there could
 > be a multiplier parameter added to SAD so that the user could set CPARM
 > and DPARM to some multiple of the off-source noise (other than 3).

    Are you running 31DEC02 with the MNJ?  If so, the task IMEAN now
returns PIXSTD to AIPS although it needs you to do an IMSTAT before
hand (or to set PIXSTD=0 may work).  You also may need to say COMPRESS
to get the new OUTPUTS verb.

CUTOFF is not used by 31DEC02 SAD.  Instead, a list of cutoffs is used
from CPARM.  If you are setting CUTOFF rather than CPARM(1) and (2)
etc then I can see why you think it is fixed at 3*rms.  The use of
CPARM is to allow the subtraction of string components first before
grubbing in the noise.  I found that a low initial cutoff would
confuse the fits of strong sources due to nearby noise bumps.

 > 
 > Task:  MFPRT
 > 
 > I modified a private version of MFPRT so that it prints out the error 
 > estimates
 > in the MF table generated by SAD.   It would be helpful if MFPRT had a
 > parameter similar to BOX in PRTAB which allows the user to choose the
 > columns to print out.
 >  

MFPRT is an odd task initially written to feed data to some special
purpose (non-official, probably non-AIPS program).  I have overloaded
that with the capability to write source lists for use by BOXES,
SETFC, etc.  I suppose we could add other overloads - what did you
have in mind?

Eric